A Scenic Area Landscape Planning Method Based on Intelligent Optimization Algorithms

Scenic area landscape planning is a complex task that requires balancing environmental preservation, aesthetic value, and visitor experience while optimizing land use efficiency. Traditional planning methods often face challenges such as dealing with high-dimensional spatial data, incorporating multiple conflicting objectives, and adapting to dynamic environmental and human factors. To address these issues, this research proposes a novel approach that integrates intelligent optimization algorithms with Variational Autoencoders (VAEs) for effective landscape planning. The intelligent optimization algorithms are employed to search for globally optimal planning layouts, while the VAEs are used for dimensionality reduction, feature extraction, and generative modeling of complex spatial patterns. The proposed method enables efficient handling of large-scale geospatial datasets and enhances the ability to simulate realistic and sustainable landscape layouts. In this study, datasets such as the USGS National Map, OpenStreetMap GIS data, and Google Earth Pro imagery were utilized to capture terrain, vegetation distribution, and infrastructure information. The VAEs learn latent representations of the landscape data, which are then optimized using algorithm Grey Wolf Optimizer (GWO) to generate planning layouts that maximize scenic quality and ecological balance. Experimental results demonstrate that the proposed approach achieves a 97.3 % accuracy and enhances visual appeal scores compared to traditional planning models. This research highlights the potential of combining deep generative models with intelligent optimization to revolutionize sustainable and adaptive scenic area planning.